SIT for the Malaria Vector Anopheles arabi - Springer Link

3 downloads 55 Views 878KB Size Report
2Department of Ancient Egypt and Sudan, The British Museum, Great .... An. arabiensis reinvasion of Egypt. The ..... the Funj Sultanate as far north as the Third.
SIT for the Malaria Vector Anopheles arabiensis in Northern State, Sudan: an Historical Review of the Field Site C. A. MALCOLM1, D. A. WELSBY2 and B. B. EL SAYED3

1School

of Biological Sciences, Queen Mary, University of London, Mile End Road, London E1 4NS, UK 2Department of Ancient Egypt and Sudan, The British Museum, Great Russell Street, London WC1B 3DG, UK 3Tropical Medicine Research Institute, National Centre for Research, PO Box 1304, Khartoum, Sudan ABSTRACT It is planned to use the sterile insect technique (SIT) as part of an area-wide integrated pest management programme to drive Anopheles arabiensis Patton, the major vector of malaria in Sudan, back from the northern-most edge of its distribution on the Dongola and Abri-Delgo Reaches of the Nile, close to the border with Egypt. The coincidental location of the field site with Upper and Middle Nubia provides a wealth of historical information that can, in lieu of direct data, reveal clues to past changes in vector distribution. Major environmental transitions have occurred across the region since the last glacial maximum, 18 000 years ago: from hyper-arid desert to tropical grassland, then to semi-desert and back to tropical desert today. Large wild animals as diverse as giraffe and hippopotamus emerged and receded. In parallel, human activities and settlement patterns changed markedly with the rise and fall, and rise again, of the Kingdom of Kush during the Kerma and Kushite periods. These factors will have facilitated the spread of mosquito populations and then, by 1500 years ago, contributed to their reduction, or demise. The "saqia" water wheel introduced at the start of medieval times brought an expansion of the human population along the Nile and presumably a gradual resurgence in mosquitoes, which continued with occasional setbacks to the present day. Isolation was almost complete except for limited dispersal downriver via the Abu Hamed Reach. Evidence of malaria in ancient times has been found in Egypt and neighbouring River Nile State, but as yet historical indicators in Northern State are limited to the accounts of foreign visitors and date from 1813. Entomological records are available from about 1908. From these it appears that An. arabiensis is the only anopheline to have been found between the Second and Fifth Cataracts and that it has remained limited to the south of Wadi Halfa over the last century with only intermittent forays into Egypt, where it caused at least two serious malaria outbreaks. KEY WORDS

Anopheles arabiensis, Northern State Sudan, River Nile, mosquito distribution, Gambiae Control Project

1. Introduction

Amongst the many criteria for judging a field site to be suitable for the control of mosquitoes using an area-wide integrated pest manamgent (AW-IPM) approach with an sterile insect technique (SIT) component, it is most desirable to find a single vector population that is isolated, unstructured, relatively small

and at low density. Today’s technology greatly facilitates the comprehensive and rapid assessment of the target population, even if starting from very little information, but the interpretation of results is not always clearcut. An historical perspective can help corroborate present-day findings, particularly in relation to the likely origin and age of the population and its history of reproductive isola-

361

M.J.B. Vreysen, A.S. Robinson and J. Hendrichs (eds.), Area-Wide Control of Insect Pests, 361–372. © 2007 IAEA.

362

C. A. MALCOLM ET AL.

Figure 1. Geology of Northern State, Sudan. tion. Other criteria may be judged with more certainty from contemporary data such as the evidence of actual or potential disease transmission and a favourable benefit/cost analysis for AW-IPM with an SIT component versus alternatives. However, even the assessment of these can be influenced by knowledge of past events, particularly since local, national and regional needs must be considered. This paper examines historical data relevant to current research to develop the SIT for control of the malaria vector Anopheles arabiensis Patton in Northern State, Sudan. The field site is primarily the Dongola Reach of the Nile, which extends down river from the Fourth to the Third Cataract (Fig. 1). An. arabiensis is restricted to within a few kilometres of the river by desert and its survival is closely asso-

ciated with the human population. Northern State encompasses Upper and Middle Nubia, the location of the Kushite civilisation that rivalled, and for at least 60 years, ruled Pharaonic Egypt. It has a very rich and ancient history with many important archaeological sites dating back before the Neolithic (7000-5000 years ago) and extending up to and including the medieval Christian and Islamic periods. This has generated information on human population movement, settlement patterns, cycles of growth and decline, land use and health, which in addition to data on climate and the environment can be used to delimit an early history of the vector population. In the nineteenth century more directly relevant information can be gleaned from the

HISTORY OF THE ANOPHELES SIT FIELD SITE IN SUDAN

accounts of travellers attracted to the area by an interest in the archaeological sites and from records of military activities (Lewis 1948). In the early twentieth century there are entomological and epidemiological records from the Wellcome Tropical Research Laboratories in Khartoum and later the Sudan Medical Service. Also well documented was the invasion of An. arabiensis into Egypt in the early 1940s, which caused a major outbreak of malaria along about 850 kilometres of the Nile (Shousha 1948) and the building of the Aswan dam which dramatically changed the landscape at the border between Northern State and Egypt. These events prompted the creation of the Gambiae Control Project, a joint protocol between the Egyptian and Sudanese Ministries of Health to prevent an An. arabiensis reinvasion of Egypt. The records of their survey and control activities, along with one study in 1986 (Dukeen and Omer 1986), are about the only entomological data from Northern State in the last 50 years, so current studies are essentially starting anew. This paper attempts to provide an overview of relevant historical data with a few highly speculative inferences on mosquito distribution. More detailed analysis, which may be possible for certain areas and time periods, is outside the intended scope and should await the outcome of current work on both archaeology and vector biology. A further caveat is that historical and archaeological studies are also limited by data paucity and subject to multiple, perhaps controversial, interpretations.

2. Present and Future

The Republic of Sudan is the largest country in Africa, covering 2.5 million square kilometres. Northern State is located in the northwestern corner, bordered by Libya and Egypt, occupying nearly one fifth of the country. It is very arid, almost entirely desert, rainfall is rare and the temperature is high throughout the year, reaching 47ºC in summer months, but can get as low as 7ºC at night in the winter.

363

As it travels north along the Dongola Reach, the Nile passes over sandstone and is flanked by wide alluvial flood plains that are suitable for cultivation of crops. As all agriculture is based on irrigation and the annual flood, most of the population (around 567 000) live alongside this stretch of the river, particularly within Dongola and Merowe provinces. Above the Fourth Cataract and below the Third Cataract, the bedrock is crystalline basement; there is little alluvial soil (Stern and Abdelsalam 1996), and therefore less scope for agriculture and a lower population density (Fig. 1). Malaria in Northern State accounts for about one third of all hospital admissions. An. arabiensis is the only vector and apparently the only anopheline. It is a member of the Anopheles gambiae Giles species complex and all but relatively recent records refer to An. gambiae. It is, however, safe to assume that these were An. arabiensis, since An. gambiae sensu stricto is only found in the far south of the country. The vector can be found all year in Dongola and Merowe Provinces, but its population density changes temporally in relation to the Nile flood and ambient temperature. Its distribution appears to be highly correlated to human presence and it is therefore more restricted along the Abu Hamed Reach and below the Third Cataract than along the Dongola Reach. The Baiyuda desert prevents immigration from the south and so, apart from human-assisted passive transport, the only route for this mosquito into Northern State is downriver via the difficult terrain above the Fourth Cataract. Just downriver from the Fourth Cataract, a dam is being built across the island of Merowe (Mirowy) to provide hydroelectric power and should be fully operational by 2008. An area of 711 square kilometres is expected to be under water creating a lake extending along the western Abu Hamed Reach. The lake shore will be rocky desert terrain unsuitable for agriculture and human habitation. It is therefore likely to pose a significant barrier to mosquito migration or dispersal downriver and should complete the isolation of the

364

C. A. MALCOLM ET AL.

Northern State vector population. However, after the dam is completed there are plans to create large irrigation channels running in parallel with the Nile all the way from Merowe to Dongola. These projects are going to dramatically transform Northern State and while the benefits are clear there is a risk they will exacerbate the malaria situation.

3. 18 000-3500 Years Ago

Given the extreme environment and near absence of human population, it is highly improbable that mosquitoes were present along the Dongola Reach 18 000 years ago, but the environment will have become progressively more suitable up to 10 000 years later. A close association with humans, such as found today, would have facilitated the presence of the vector on the Dongola Reach between 7000 and 3500 years ago. Up until the second millennium BC the population was probably not isolated and an origin in the region of the south-western frontier with present day Chad was at least as probable as one in central Sudan. By the first millennium BC the environment was becoming more similar to today, the human population had declined and it is likely that mosquito populations were not sustained. Thereafter recolonization would have to have been downriver via the Abu Hamed Reach. The earliest known presence of humans in the north-western part of Sudan dates back at least 300 000 years, however by 18 000 years ago, the whole area was hyper-arid desert (Petit-Maire et al. 2000, Hoelzmann et al. 2001, Adams 2002) with presumably only a small population along the river margins (Fig. 2). The flow of the Nile was substantially less than today and seasonally may have been completely dry in places. As the desert receded, small pockets of humans were living close to the Second Cataract between 15 000 and 9000 years ago (Garcea 2004). About 3000 years later almost all of Northern State was tropical grassland. The Sahara started to expand again, and by the later Holocene, tropical semi-desert was encroaching on the

Dongola Reach (Fig. 2) (Petit-Maire et al. 2000, Hoelzmann et al. 2001, Adams 2002). Hunter gatherers started to change to a more pastoral way of life and livestock herding had appeared by the start of the Neolithic (7000 years ago). In the northern Dongola Reach there was considerable occupation along the banks of the Nile palaeochannels that flowed to the east of the present day river, close to the Kerma and Seleim Basins and the Wadi-alKhowi. These sites followed a loose settlement pattern across a wide area, rather than discrete densely populated centres, which developed and intensified over the next 2000 years (Fig. 3). One well-preserved settlement near Kerma was dated at 6500 years ago (Honegger 2001). Domestic livestock first appeared in the Wadi Howar around 4200 BC. Over the next 1000 years the Lower Wadi Howar was to remain fairly densely populated as settlement to the west increased, in particular along the Middle Wadi Hower and Djebel Tageru and in the West Nubian Palaeolake basin, allowing links between the Nile valley and Chad basin (Hoelzmann et al. 2001, Edwards 2004). Environmental conditions will almost certainly have permitted migration of mosquitoes including An. arabiensis to the Dongola Reach and to Egypt before this time, but the more contiguous series of human settlements along the Wadi Howar and up to Kerma will have provided an even more favourable route. The source of the Wadi Howar (also known as the Yellow Nile) was in the border region of present day Chad. This could have been the origin of an early Dongola Reach An. arabiensis population and even of some present day mosquito populations in Egypt. Much less is known of human settlements along the Wadi el-Melik, the southern Dongola Reach and the Abu Hamed Reach, although work in the latter two areas is ongoing, so alternative or additional migration routes cannot yet be excluded. By about 4200 years ago the lower Wadi Howar was drying up and the focus of settlement moved westwards. Cultural divergence with the northern Dongola Reach population

HISTORY OF THE ANOPHELES SIT FIELD SITE IN SUDAN

365

Figure 2. Changing vegetation patterns across northern Sudan from 18 000 years ago to the present day. increased, indicating a marked reduction in contact (Jesse 2004). The latter was growing, primarily around a more urbanized centre at Kerma and by 4000 years ago had become the substantial Kingdom of Kush. This was probably the earliest major kingdom in subSaharan Africa and extended along at least 700 kilometres of the Nile to encompass both Middle and Lower Nubia. The era is known as the Kerma Period and lasted over 1000 years. The settlement patterns while still extended, became denser and mostly followed the Hawawiya, Alfreda and Seleim Nile paleochannels (Welsby 2001) (Fig. 3). The envi-

ronment was mainly lightly wooded savannah; intensive agriculture had increased and there was a strong emphasis on livestock management. A wide range of large wild animals was still present including giraffe, hippopotamus, lions and antelope (Edwards 2004). Sizable mosquito populations could easily have been sustained at this time, but probably relatively isolated from populations outside the region as the surrounding area was becoming semi-desert (Fig. 2) and the wadis to the south were no longer flowing into the Nile. The only available migration or dispersal route for mos-

366

C. A. MALCOLM ET AL.

Figure 3. Settlement patterns in northern Sudan during the Kerma and Kushite Periods (adapted from Welsby 2001 and Edwards 2004). quitoes from this time onwards will have been down the river Nile. There was human settlement further south in the Letti Basin and in the vicinity of the Fourth Cataract. As mentioned before, most of the Abu Hamed Reach appears inhospitable, particularly at the Fourth Cataract end, but it was nevertheless densely occupied at various times, including during the Kerma Period. Without mechanical devices to assist irrigation, the seasonally dry riverbeds between the numerous islands actually made it a more suitable location for cultivating crops than some ostensibly richer areas further downriver. The rocky terrain will also have

provided protection.

4. 3500-1500 Years Ago

By about 3500 years ago the Egyptians had conquered Kerma state, with the destruction of Kerma itself and were to dominate the area for over 400 years. The expansion of the desert was by now approaching its present day boundary. Where they have recently been studied in detail, settlements adjacent to the Seleim Basin and Wadi-al-Khowi were found to have been disappearing as the palaeochannels dried up, with new sites appearing along

HISTORY OF THE ANOPHELES SIT FIELD SITE IN SUDAN

the bank of the present day Nile channel. The major population centre became the Pharaonic site at Kawa. Much less is known about the rural population in the period under Egyptian control and for several centuries after, but clearly the population diminished substantially and the settlement pattern changed to a small number of disconnected population centres. This pattern is still evident during the Kushite revival, which was to follow in the first millennium BC (Fig. 3) (Welsby 2001). Lower Nubia, which has been more extensively studied, had become largely depopulated and was not farmed for about 1000 years (Edwards 2004). Therefore by 2800 years ago, with a much reduced human population, located in pockets perhaps tens of kilometres apart, few if any large wild animals and a semidesert environment, mosquito populations would have been fractured and greatly diminished, perhaps absent altogether. After the collapse of the Egyptian New Kingdom, the Kingdom of Kush re-emerged to form an empire that survived for over 1000 years, including over half a century when Kushite Kings conquered and ruled Egypt as the 25th Dynasty. The Kushite Period can be divided into the Napatan and the Meroitic depending on the location of political power. Napata lies downriver from the Fourth Cataract, encompassing Jebel Barkal, Sanam and Abu Dom (present day Merowe and Kareima). Meroe lies between Shendi and Atbara, to the north of Khartoum. On the Dongola Reach other important sites for Napatan occupation were in the Letti Basin, Kerma, Tabo on Argo island and Kawa probably the biggest covering 40 hectares (estimated 6200-7800 people) (Fig. 3) (Edwards 2004, Welsby 2004). Below the Third Cataract there was a thin scatter of settlements, which were probably short-term and mainly existed to maintain links with Egypt (Edwards 2004). Therefore, despite a probable increase in the size of the human population it was still mainly concentrated in specific sites located far apart. With the shift of power to Meroe in the later first millennium BC many Napatan set-

367

tlements were eclipsed. Jebel Barkel remained of religious importance, but other sites contracted or disappeared. In the Northern Dongola Reach the main focus may have moved back to Kerma. Archaeological sites in the desert indicate that links with Meroe were maintained across the Baiyuda rather than upriver. Throughout the Kushite Period it seems unlikely that conditions had become any more favourable for mosquitoes along the Dongola Reach. The increasingly harsh environment will only have been partly mitigated by human activities, since despite a period of increased size the population was divided into isolated pockets. This situation probably remained into the middle of the first millennium AD only starting to change as human settlement patterns altered with the introduction of the saqia water wheel (Edwards 2004). The end of the Kushite civilization brought a transition to two new kingdoms along the Nile in Nubia during the period 1650 to 1450 years ago: Nobadia, extending from the First to the Third Cataract, and Makuria on the Dongola Reach. The major urban centres at Kawa and Kerma had disappeared, but there was significant occupation in the Letti Basin, and especially at Old Dongola, which was to become the capital of medieval Makuria. This period also saw an expansion of settlement in the Fourth Cataract and further upstream (Edwards 2004).

5. 1500-200 Years Ago

Human population density and settlement patterns along the Dongola Reach changed dramatically in medieval times. A major factor was the introduction of the saqia water wheel, which along with changing farming practices and new multi-harvest crops contributed to the expansion of the population into areas that had been largely uninhabitable since the Kerma Period. The Kingdom of Nobadia controlled Lower and Middle Nubia between the Third and First Cataracts with a population in the order of 4000 divided into about 40 settlements, with perhaps as many as 20 people per square kilometre towards the south end of the

368

C. A. MALCOLM ET AL.

Batn-el-Hajar (Edwards 2004). The much better agricultural resources along the Dongola Reach will have sustained a much larger population. The Makuria Kingdom’s main population centre lay at Old Dongola. Further north, settlements developed the west bank, where there had been limited human activity previously. This new settlement pattern was becoming more and more like the continuous settlement that exists along both banks today. There was also settlement along the western part of the Abu Hamed Reach above the Fourth Cataract. The occupation of the Nile banks expanded and by the early eighth century the two kingdoms unified under the King of Makuria (Edwards 2004). From about 1500 years ago, humans were progressively creating conditions that were likely to facilitate expansion of mosquito populations along the length of the Dongola Reach, whereas for the previous 1000 years it’s likely that any population would have been small and isolated, with little or no scope for recolonization from neighbouring human settlements or from outside the region. It would seem more likely that a species such as An. arabiensis, which is less well adapted to survival alongside humans than for example Culex quinquefasciatus Say, would have disappeared altogether. For perhaps the past 1700 years the only route into the Dongola Reach for mosquitoes was downriver and this remains the situation today, with the possible exception of passive transport facilitated by the railway to Wadi Halfa and Kareima, completed in the early twentieth century, and modern day transport. In medieval times, the western Abu Hamed Reach was extensively occupied, which would have increased the likelihood of mosquito reinvasion downriver. Therefore whether or not An. arabiensis had disappeared previously, it was almost certainly present along the Dongola Reach for most of the last 1500 years. The biggest setbacks for both humans and mosquitoes during this time were mainly associated with floods and droughts, although major conflicts and social change were still to occur. The environmental history

of the area is relatively poor for the first millennium AD, but there are records of several periods of a decade or so, when the Nile level was very low, other periods of high Nile (Adams 2001), and some years when ice formed on the Nile in Egypt, particularly in the ninth and tenth centuries. From 1300 to 1522 the Nile floods were generally good. After this and up to part way through the eighteenth century Nile levels and humidity were higher, but there were some droughts affecting the whole of Sudan in the late seventeenth century. In the eighteenth and nineteenth centuries the climate in general became more arid bringing more droughts, but interspersed with some heavy and destructive floods, with a very wet period in the 1790s (Nicholson 1978, Edwards 2004). The data available over the last 500 years is quite detailed and could be used to make a more in-depth analysis of the likely impact on mosquito populations than the superficial account presented here. Nevertheless, it is clear that as with any mosquito at the edge of its species distribution, numbers will have fluctuated dramatically and the population structure will have been disrupted with successive contractions and expansions along parts of the Nile within Northern State, but it does not appear likely to have been eradicated from the area altogether at any point. The human population will have endured famine and conflict associated with major floods and droughts, and other major upheavals caused by political changes such as the expansion of the Funj Sultanate as far north as the Third Cataract and the expansion of the Ottoman Empire south (Edwards 2004). Nevertheless, the contiguous human settlement pattern along the Dongola Reach was maintained and became progressively more continuous, facilitating a more extended and presumably more homogenous mosquito population. A more difficult question is the extent to which the Dongola Reach An. arabiensis population was isolated. The key is the Fourth Cataract and the western end of the Abu Hamed Reach, which was densely populated during the Kerma Period around 2500-1500

HISTORY OF THE ANOPHELES SIT FIELD SITE IN SUDAN

BC, and again in the post-Meroitic and medieval periods, between AD 350-1500. The population levels in the period in between are uncertain as work in the area is still underway. Continuity of occupation up to the present day is probable. The absence of human settlement would be a strong indication of a period of isolation. Human occupation does not however necessarily mean that mosquito migration was possible, since amongst other factors, the nature and size of the settlement and the prevailing wind and Nile level will all affect the likelihood. This again merits a more detailed analysis, but will have to await the results from the current Merowe Dam archaeological salvage project.

6. Last 200 Years

Malaria in Egypt has been confirmed from over 5000 years ago and is thought to have been widespread (Cerutti et al. 1999). Direct evidence for malaria in Northern State before the last 200 years does not yet exist. In neighbouring River Nile State, a Kushite, postMeroitic and medieval cemetery at Gabati near Atbara, shows mortality patterns associated with anaemia that suggest malaria was present there (Judd 2005). Similar studies are in progress within Northern State. Lewis (1948) provided evidence for malaria-like illness in Dongola in the early nineteenth century gleaned from the literary accounts of travellers to the area. The earliest was 1813 described by Burkhardt (1819) in the Dongola area of a fever which occurred in epidemics, but not every year and which was often fatal. Entomological records confirming the presence of An. arabiensis first became available after the founding of the Wellcome Tropical Research Laboratories at Gordon Memorial College, Khartoum, in 1902 under the first director Dr Andrew Balfour. In the beginning medical, entomological and sanitation work in Khartoum took priority and then expeditions to the south. Many of the difficulties encountered in this early field work were overcome in 1907, by outfitting a boat as a floating laboratory (D’Arcy 1999), and this approach has

369

also been adopted by the Gambiae Control Project. King (1908) was perhaps the first to record An. arabiensis in Northern State in 1906 and 1907. From around 1918, outbreaks of malaria in Egypt at Nag Hammadi, Armant and KomOmbo were associated with the introduction of sugarcane. Few deaths were recorded and the incidence progressively declined with little or no intervention. This could have been due to Anopheles pharoensis Theobald, but a malaria epidemic in 1919-20 was thought more likely to be due to An. arabiensis. This followed heavy rains in the area around Ed Derr about 80 kilometres into Egypt (Fig. 4) and resulted in about 1000 deaths. The most detailed documented invasion of Egypt by the vector started in 1942 reaching Asyut, 850 kilometres downriver from the border. It caused a malaria epidemic estimated to have resulted in over 10 000 deaths within two years. An extensive and well-executed control programme successfully eradicated it from Egypt by early 1945 (Shousha 1948). Further control activities removed it from the Wadi Halfa area between Saras and Faras by early 1996 (Lewis 1948). An. arabiensis was not controlled again in the vicinity of Wadi Halfa or further south until 1951, and was not found north of Ferka (120 kilometres south from the border) until late 1950, when it appeared at Aneiba in Egypt and then in neighbouring villages. It was again controlled with DDT and larviciding oil. Thereafter, it was decided to maintain a control programme using oil between Saras and Aneiba. In 1954, the Sudanese and Egyptian Ministries of Health agreed a joint programme to control An. arabiensis south of Wadi Halfa (Shousha 1948, Lewis 1956). This was to be the forerunner to the Gambiae Control Project, which started in 1970. The objectives of this control programme were to monitor and maintain an An. arabiensis-free zone, the “Red Zone”, from Aswan in Egypt upriver to a point well inside Sudan and through control activities further upriver (the “Yellow” and “Green” Zones) to try and extend it. In 1970 the northern-most limit of

370

C. A. MALCOLM ET AL.

Figure 4. Northern distribution of Anopheles arabiensis over the last 100 years. An. arabiensis was judged to be Akasha and by 1996 this had been pushed back close to Abri (Fig. 4). All historical entomological records for Northern State indicate An. arabiensis is the only anopheline, and therefore the only malaria vector present, with the exception of reports of An. pharoensis just south of Wadi Halfa at Abka and Anopheles multicolor Cambouliu and Anopheles d’thali Patton a little further north at Faras. Two others, Anopheles sergen-

tii (Theobald) and Anopheles coustani Laveran merit attention since they have come close to the border. An interesting observation is that four of these species are found in other parts of Sudan; An. pharoensis in particular is relatively common south of Khartoum (Shousha 1948, Lewis 1956, Cope et al. 1995). If these species are indeed absent from Northern State, or present only periodically in small numbers, it would be of interest to determine the degree of genetic differentiation

HISTORY OF THE ANOPHELES SIT FIELD SITE IN SUDAN

between the Egyptian and Sudanese populations, since it is conceivable that their separation dates back to the Kerma Period or before, when the environment was much more favourable for migration along the Nile. Two important components of research and development for the use of the SIT are population studies on the target and an understanding of its relationship with other species likely to be affected by its suppression or eradication. Other potential vectors are the most immediate concern. The fact that they appear to be absent is a major advantage to the project, but there is clearly some merit in knowing why they are absent and how easily that situation might change. Present day population studies increasingly rely on the use of molecular markers, but as with most approaches the data can often have multiple interpretations. Preliminary studies based on analysis of microsatellite DNA data suggest that the Northern State An. arabiensis population is isolated by the desert and partly by the western stretch of the Abu Hamed Reach. Analysis of mitochondrial DNA, however gives a different picture. This is easily resolved if it is assumed that the isolation is sufficiently recent to only show up with the more rapidly-evolving microsatellite DNA. This interpretation appears therefore to be consistent with the inferences made from the historical data. Despite being highly speculative, the scenario proposed for past changes in the vector population, does provide an interesting perspective that may prove more useful as further analysis of the present day population becomes available and as archaeological studies continue in Northern State, particularly at the present time around the Fourth Cataract.

7. Conclusions

It is suggested that An. arabiensis was absent from Northern State 18 000 years ago, but between 8000 and 3500 years ago it migrated from any of several diverse locations to establish a presence in the Dongola Reach and spread further into Egypt. It then receded from

371

Egypt and most probably also the Dongola Reach. Other mosquitoes found today in both countries may also have receded from the Northern State area and some, such as An. pharoensis, may never have re-established a significant presence. Within the last 1500 years An. arabiensis became established again in Northern State, almost certainly from upriver, and despite many bottlenecks maintained a continuous presence reaching as far as the Second Cataract, with only intermittent incursions into Egypt. A more detailed investigation may throw more light on the factors, frequency and duration of likely bottlenecks during the last 1500 years and the extent to which the population has been supplemented by migration or dispersal downriver along the Abu Hamed Reach.

8. References

Adams, J. M. (ed.). 2002. Global land environments since the last interglacial. http://members.cox.net/quaternary/ Adams, W. Y. 2001. Meinarti II. The early and classic Christian phases. Archaeopress, Oxford, UK. Burkhardt, J. L. 1819. Travels in Nubia. eBooks@Adelaide 2004. http://etext.library. adelaide.edu.au/b/burckhardt/john_lewis//nu bia/ Cerutti, N., A. Marin, E. R. Massa, and D. Savoia. 1999. Immunological investigation of malaria and new perspectives in paleopathological studies. Bollettino-Societa Italiana Biologia Sperimentale (Napoli)/ Bulletin of the Italian Society of Biology 75: 17-20. Cope, S. E., A. M. Gad, and S. M. Presley. 1995. New record of the malaria vector Anopheles sergentii in the southern Nile Valley of Egypt. Journal of the American Mosquito Control Association 11: 145-146. D’Arcy, P. F. 1999. Laboratory on the Nile: a history of the Wellcome Tropical Research Laboratory. Pharmaceutical Products Press, Haworth Press Inc., Binghamton, NY, USA. Dukeen, M. Y. H., and S. M. Omer. 1986. Ecology of the malaria vector Anopheles

372

C. A. MALCOLM ET AL.

arabiensis Patton (Diptera: Culicidae) by the Nile in northern Sudan. Bulletin of Entomological Research 76: 451-467. Edwards, D. N. 2004. The Nubian past: an archaeology of the Sudan. Routledge, London and New York, UK and USA. Garcea, E. 2004. The Palaeolithic and Mesolithic, pp. 20-24. In Welsby, D. A., and J. R. Anderson (eds.), Sudan ancient treasures. The British Museum, London, UK. Hoelzmann, P., B. Keding, H. Berke, S. Kropelin, and H-J. Kruse. 2001. Environmental change and archaeology: lake evolution and human occupation in the Eastern Sahara during the Holocene. Palaeogeography, Palaeoclimatology, Palaeoecology 169: 193-217. Honegger, M. 2001. Evolution de la société dans le bassin de Kerma (Soudan) des derniers chasseurs-cueilleurs au premier royaume de Nubie. Bulletin de la Société Française d’ Égyptologie 152: 12-27. Jesse, F. 2004. The Wadi Howar, pp. 53-60. In Welsby, D. A., and J. R. Anderson (eds.), Sudan ancient treasures. The British Museum, London, UK. Judd, M. 2005. Gabati: health in transition. Sudan and Nubia 8: 84-89. King, H. H. 1908. Report on economic entomology. Third Report Wellcome Tropical Medicine Research Laboratories, Khartoum, Sudan. Lewis, D. J. 1948. Early references to malaria

near Dongola. Sudan Record and Notes 29: 218-220. Lewis, D. J. 1956. The anopheline mosquitoes of the Sudan. Bulletin of Entomological Research 47: 475-494. Nicholson, S. 1978. Climatic variations in the Sahel and other African regions during the past five centuries. Journal of Arid Environments 1: 3-24. Petit-Maire, N., P. Bouysse, J. L. de Beaulieu, G. Boulton, P. Kershaw, O. Litsitsyna, T. Partrige, U. Pflaumann, H. Schultz, J. Soons, B. van Vliet-Lanoe, and G. Zhengtang. 2000. Geological records of the recent past, a key to the near future world environments. Episodes 23: 230-246. Shousha, A. T. 1948. Species eradication. The eradication of Anopheles gambiae from upper Egypt 1942-1945. Bulletin of the World Health Organization 1: 309-352. Stern, R. J., and M. G. Abdelsalam. 1996. The origin of the Great Bend of the Nile from SIRC/XSAR imagery. Science 274: 1696-1698. Welsby, D. A. 2001. Life on the desert edge. Seven thousand years of settlement in the Northern Dongola Reach of the Nile. The British Museum, London, UK. Welsby, D. A. 2004. Kawa, pp. 148-157. In Welsby, D. A., and J. R. Anderson (eds.), Sudan ancient treasures. The British Museum, London, UK.